Hi,
I've been following the discussion about VGA emulation and Bochs-RFB.
I like the VNC approach very much, and so I'm a fan of a Bochs-RFB
implementation.
I've read a lot on the VNC pages, and there seems to be a GGI VNC
target available too, although it may be somewhat outdated.
If we can get Plex86 (and/or Bochs) to run on top of GGI (KGI) then
we'd have everything covered:
1) We can run Bochs/Plex fullscreen without X
2) We can use the GGI X compatability to run under X
3) We can use the GGI VNC target to export the GUI over the net
4) If we start with a Linux guest driver communicating with a Linux
host, then we can more or less directly route the GGI/KGI calls
from the guest to the target. (you could see this as adding a
new 'plex86' target to GGI). This way the guest graphics could be
almost as fast as the host graphics.
I haven't put much time into studying the GGI/KGI implementations,
or how difficult it would be to implement this kind of architecture, but
it seems to me that this could be the right way to go.
If GGI/KGI is not considered a good option, we could restrict ourselves
to the Linux framebuffer device (fbdev), but then there's no support
for VNC.
For more information on GGI/KGI see this page:
http://www.ggi-project.org
Of course I agree that we need to keep a basic VGA16 hardware
compatibility layer for all the OS's for which no accelerated drivers
are available yet.
Well, these are just my thoughts, what do you guys think?
> Just for my own curiosity I would like to know how many people are
> interested in the Bochs-RFB project. If you think this idea is useful
> please respond to this message with a brief comment.
Sorry, this wasn't very brief :-), but I like your idea!
Mark Hurenkamp
